I have noticed that when using double-clustering there is a statement that only reports the number of the second type of clusters. Example: imagine we have 100 firms and 10 years per firm. Then:
reghdfe depvar indepvar, absorb(firm year) cluster(firm)
it says "Std. err. adjusted for 100 clusters in firm"
reghdfe depvar indepvar, absorb(firm year) cluster(year)
it says "Std. err. adjusted for 10 clusters in year"
reghdfe depvar indepvar, absorb(firm year) cluster(firm year)
it says "Std. err. adjusted for 10 clusters in firm year"
NOTE: the correct number of clusters is reported above the main output, the bug is only in that statement, for example:
Number of clusters (firm) = 100 Number of clusters (year) = 10
Bug report
I have noticed that when using double-clustering there is a statement that only reports the number of the second type of clusters. Example: imagine we have 100 firms and 10 years per firm. Then:
reghdfe depvar indepvar, absorb(firm year) cluster(firm) it says "Std. err. adjusted for 100 clusters in firm" reghdfe depvar indepvar, absorb(firm year) cluster(year) it says "Std. err. adjusted for 10 clusters in year" reghdfe depvar indepvar, absorb(firm year) cluster(firm year) it says "Std. err. adjusted for 10 clusters in firm year"
NOTE: the correct number of clusters is reported above the main output, the bug is only in that statement, for example:
Number of clusters (firm) = 100
Number of clusters (year) = 10
(Std. err. adjusted for 10 clusters in firm year)